NFL Rostergen: R-Shiny Application for Optimizing NFL Team Construction Using Genetic Algorithms

Abstract

Roster construction is critical in any sport for the team's subsequent performance. The significance of this project is to create better team construction strategies for American Football teams in the NFL, and the solution can be extrapolated to other sports. Multiple factors influence the selection process, such as current team construction, player skill, financial constraints, which may complicate the process of selecting 53-man rosters since there are several permutations for a greedy-search heuristic. The goal of this project is to use a Genetic Algorithm to objectively select players by integrating an ML-based fitness function used to evaluate a team’s quality.

Keywords: NFL, Roster-formulation, Genetic-Algorithm, Fitness-Function